Cases undergoing cardiac surgery will be studied with a behavioral methodology pre and postoperatively for the occurrence of deficits in higher cerebral function. These latter should largely reflect the predominant pathologic involvement of the parietooccipital regions. The behavioral methods involve matching-to-sample, oral and written naming of stimuli presented in visual, palpated, and auditory form. Special emphasis will be given to those materials presentable in all three sensory modalities (i.e., a police whistle as a picture, sounds, and manipulable object). Visual stimuli will be divided into non-forms and forms, the non-forms including color; the forms will be subdivided into animated and unanimated stimuli, the former including body parts and positions; the unanimated stimuli will be subdivided into manipulable and ordinarily non-manipulable objects; the latter will be subdivided into pictures of large stimuli, letters, words, digits, and simple forms. In addition, patients will be trained on special gradient discrimination tests preoperatively for postoperative retention testing. Pathological material will be studied by serial section techniques for topography of infarct and for clinico-pathologic correlation. All subjects will be given pre and postoperative neurologic, formal visual field, and electro-encephalogram examinations. The variety of deficits will be delineated, those of transient nature being separated from those permanent, and efforts made in pathologic and case history analysis to determine avoidable causal factors.